154 Dr. J. Kerr's Electro-optic 



important piece in the first of my two papers already referred 

 to, I think that here I need only reproduce the diagram. 



Optical Arrangements. — The plate cell, charged nearly full 

 with the dielectric liquid, is stationed between the polarizing 

 and analyzing Nicols, the faces of the liquid plate being 

 always perpendicular to the line of vision. In most cases, the 

 object viewed in the polariscope is a narrow and sharply-out- 

 lined flame, a flat and thin paraffin-flame presented edgeways, 

 and seen exactly through the centre of the electric field. The 

 first Nicol is fixed with its principal section at 45° to the 

 horizon, or to the lines of force; and the second Nicol is fixed 

 at extinction. When the light is well restored by electric 

 force the character of the effect as positive or negative is 

 generally detected at a glance, and with perfect certainty, by 

 the use of the hand-compensator. This is a small rectangular 

 slip of thin plate glass, held in the observer's hands between 

 the two Nicols, and simply stretched or compressed in a direc- 

 tion parallel to the lines of force. 



Electric Arrangements. — These vary with circumstances, and 

 are sometimes all employed successively in the examination of 

 one dielectric. The liquid may be tried (1) as a nonconductor, 

 (2) as a conductor, but without the jar, (3) with the jar, 

 (4) with intense discharges of a Ruhmkorff's coil. 



Every new liquid is tested for insulation in the first place. 

 The outer balls of the cell are connected respectively with 

 prime conductor and earth by unbroken wires, and the machine 

 is set in motion. The test consists in bringing up the knuckle 

 or an earth-connected ball to contact with prime conductor, 

 and receiving, or not receiving, a spark. 



(1) To try a liquid electro-optically as a nonconductor. 

 The cell is placed in the polariscope at extinction, its outer 

 balls being connected with prime conductor and earth by un- 

 broken wires ; the machine is set in motion, and the polari- 

 scope is watched for a restoration of the light from extinction. 

 When there is a clear effect, it is tested by the hand-compen- 

 sator, and characterized as positive or negative. This method 

 has given distinct results with every moderately good insulator 

 yet examined. 



